Squamous cell carcinomas of the head and neck (HNSCCs) remain largely poor prognosis due to late diagnosis and rapid invasion and metastasis. Mutations and loss of expression of transforming growth factor a (TGFbeta) signaling components and paradoxical overexpression of TGFbeta1 ligand frequently occur in human HNSCCs. However, the exact roles of these alterations in HNSCC carcinogenesis remain unknown. Based on our preliminary data and current knowledge, we hypothesize: 1) Loss of TGFbeta signaling components in tumor epithelia promotes HNSCC growth primarily due to the loss of the growth inhibitory effect of TGFbeta1 on tumor cells. 2) Overexpression of TGFbeta1 by tumor epithelia promotes HNSCC invasion via a TGFbetaRII or Smad(s)- independent signaling pathway or via paracrine effects on tumor stroma. Thus, HNSCCs which have increased TGFbeta1 with concurrent loss of TGFbeta signaling components in tumor epithelia will have the worst prognosis. To test our hypotheses, we will establish transgenic/knockout mouse models in which TGFbetaRII or Smad4 is deleted, or Smad7 is overexpressed in head and neck epithelia, and observe whether loss of TGFbeta signaling in the epithelia results in spontaneous HNSCC formation. To mimic an etiological risk factor for HNSCC development, chemical carcinogenesis protocols will be applied to oral cavity of these transgenic/knockout mice and observe whether loss of TGFbeta signaling in tumor epithelia confers a higher susceptibility to HNSCC formation and progression. To examine the role of TGFbeta1 ligand in HNSCC carcinogenesis and determine whether the combination of increased TGFbeta1 ligand and decreased TGFbeta signaling components in HNSCC epithelia results in rapid tumor invasion, we will generate compound mice in which TGFbeta1 can be inducibly expressed in HNSCCs with or without the deletion of TGFbetaRII or Smad4. Pathological processes related to tumor progression and metastasis, and molecular changes related to these processes, will be analyzed in the mouse models generated in this proposal. The proposed studies will provide important insights into molecular mechanisms of HNSCC development and progression, as well as aid in the identification of novel biomarkers for prognosis and therapeutic targets of HNSCCs. The animal models generated in this proposal will provide valuable resources for testing therapeutic approaches for HNSCCs in the future.